By default, yt-dlp tries to download the best available quality if you don't pass any options. This is generally
equivalent to using -f bestvideo*+bestaudio/best. However, if multiple audiostreams is enabled (--audio-multistreams),
the default format changes to -f bestvideo+bestaudio/best. Similarly, if ffmpeg is unavailable, or if you use yt-dlp
to stream to stdout (-o -), the default becomes -f best/bestvideo+bestaudio.
So I think that it should normally pull down the best audio unless you get into some situation where YouTube doesn’t offer a format that simultaneously has the combination of highest audio quality with the highest video quality; if it has to do so to get the highest video quality then, it’ll sacrifice audio quality.
EDIT: Hmm. I could have sworn that there was more text about prioritizing relative audio and video quality at one point in the man page, but I don’t see anything there now. Maybe it can just always get the best audio quality, regardless of video quality, can pull 'em entirely separately.
I’m pretty sure that it defaults to best quality.
goes looking at man page
So I think that it should normally pull down the best audio unless you get into some situation where YouTube doesn’t offer a format that simultaneously has the combination of highest audio quality with the highest video quality; if it has to do so to get the highest video quality then, it’ll sacrifice audio quality.
EDIT: Hmm. I could have sworn that there was more text about prioritizing relative audio and video quality at one point in the man page, but I don’t see anything there now. Maybe it can just always get the best audio quality, regardless of video quality, can pull 'em entirely separately.